U.S. patent number 7,292,622 [Application Number 10/680,492] was granted by the patent office on 2007-11-06 for method and apparatus for raking in a wireless network.
This patent grant is currently assigned to Freescale Semiconductor, Inc.. Invention is credited to John W. McCorkle.
United States Patent |
7,292,622 |
McCorkle |
November 6, 2007 |
**Please see images for:
( Certificate of Correction ) ** |
Method and apparatus for raking in a wireless network
Abstract
A raking receiver is provided in a wireless network. The raking
receiver includes an antenna, first through Nth wavelet forming
networks, first through Nth weighting mixers, a summer, a path
mixer, and a signal processing circuit. The antenna receives an
incoming signal. The first through Nth wavelet forming networks
produce first through Nth locally generated wavelets. The first
through Nth weighting mixers multiply the first through Nth locally
generated wavelets by first through Nth weighting values,
respectively, to produce first through Nth weighted wavelets. The
summer adds together the first through Nth weighted wavelets to
produce a weighted correlation input signal. The path mixer
multiplies the incoming signal with the weighted correlation input
signal to produce a correlated signal. And the signal processing
circuit receives the correlated signal and produces a digital bit
value. N is preferably an integer greater than 1.
Inventors: |
McCorkle; John W. (Vienna,
VA) |
Assignee: |
Freescale Semiconductor, Inc.
(Austin, TX)
|
Family
ID: |
32045401 |
Appl.
No.: |
10/680,492 |
Filed: |
October 8, 2003 |
Prior Publication Data
|
|
|
|
Document
Identifier |
Publication Date |
|
US 20040066842 A1 |
Apr 8, 2004 |
|
Related U.S. Patent Documents
|
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
Issue Date |
|
|
60416519 |
Oct 8, 2002 |
|
|
|
|
Current U.S.
Class: |
375/148; 375/150;
375/347 |
Current CPC
Class: |
H04B
1/71637 (20130101); H04B 1/719 (20130101) |
Current International
Class: |
H04B
1/707 (20060101); H04L 1/02 (20060101) |
Field of
Search: |
;375/147,148,150,346,347,349 ;370/335,342,441 ;455/65,63.1 |
References Cited
[Referenced By]
U.S. Patent Documents
Primary Examiner: Deppe; Betsy L.
Parent Case Text
CROSS-REFERENCE TO RELATED PATENT DOCUMENTS
This application relies for priority on U.S. provisional
application Ser. No. 60/416,519, by John W. McCorkle, filed Oct. 8,
2002, entitled "A METHOD OF ACQUIRING AN ULTRAWIDE BANDWIDTH
SIGNAL," the contents of which is hereby incorporated by reference
in its entirety.
The present document contains subject matter related to that
disclosed in commonly owned, co-pending application Ser. No.
09/209,460 filed Dec. 11, 1998, entitled ULTRA WIDE BANDWIDTH
SPREAD-SPECTRUM COMMUNICATIONS SYSTEM; application Ser. No.
09/972,966 filed Oct. 10, 2001, entitled ULTRA WIDE BANDWIDTH NOISE
CANCELLATION MECHANISM AND METHOD; application Ser. No. 09/685,197
filed Oct. 10, 2000, entitled MODE CONTROLLER FOR SIGNAL
ACQUISITION AND TRACKING IN AN ULTRA WIDEBAND COMMUNICATION SYSTEM;
application Ser. No. 09/684,400 filed Oct. 10, 2000, entitled ULTRA
WIDEBAND COMMUNICATION SYSTEM, METHOD, AND DEVICE WITH LOW NOISE
PULSE FORMATION; application Ser. No. 09/685,195 filed Oct. 10,
2000, entitled ULTRA WIDE BANDWIDTH SYSTEM AND METHOD FOR FAST
SYNCHRONIZATION; application Ser. No. 09/684,401 filed Oct. 10,
2000, entitled ULTRA WIDE BANDWIDTH SYSTEM AND METHOD FOR FAST
SYNCHRONIZATION USING SUB CODE SPINS; application Ser. No.
09/684,782 filed Oct. 10, 2000, entitled ULTRAWIDEBAND
COMMUNICATION SYSTEM, METHOD, AND DEVICE WITH LOW NOISE RECEPTION;
application Ser. No. 10/214,183 filed Aug. 8, 2002, entitled MODE
CONTROLLER FOR SIGNAL ACQUISITION AND TRACKING IN AN ULTRA WIDEBAND
COMMUNICATION SYSTEM; and application Ser. No. 09/685,200 filed
Oct. 10, 2000, entitled LEAKAGE NULLING RECEIVER CORRELATOR
STRUCTURE AND METHOD FOR ULTRAWIDE BANDWIDTH COMMUNICATION, the
entire contents of each of which being incorporated herein by
reference.
Claims
I claim:
1. A raking receiver in a wireless network, comprising: an antenna
for receiving an incoming signal; first through N.sup.th main path
wavelet forming networks for producing first through N.sup.th
locally generated main wavelets, respectively; first through
N.sup.th main path weighting mixers for multiplying the first
through N.sup.th locally generated main wavelets by first through
N.sup.th main weighting values, respectively, to produce first
through N.sup.th main weighted wavelets; a main path summer for
adding together the first through N.sup.th main weighted wavelets
to produce a weighted main correlation input signal; a main path
mixer for multiplying the incoming signal with the weighted main
correlation input signal to produce a main correlated signal; and a
main path signal processing circuit for receiving the main
correlated signal and producing a digital bit value, wherein N is
an integer greater than 1.
2. A raking receiver in a wireless network, as recited in claim 1,
wherein the main path signal processing circuit further comprises:
a main path filtering circuit for filtering the main correlated
signal to produce a main filtered signal.
3. A raking receiver in a wireless network, as recited in claim 2,
wherein the main path signal processing circuit further comprises:
a main path analog-to-digital converter for converting the main
filtered signal to a digital bit value.
4. A raking receiver in a wireless network, as recited in claim 2,
wherein the main path signal processing circuit further comprises:
a main path amplifier chain for amplifying the main correlated
signal before it is provided to the main path filtering
circuit.
5. A raking receiver in a wireless network, as recited in claim 1,
wherein the main path signal processing circuit further comprises:
a main path integrating circuit for integrating the main correlated
signal to produce a main integrated signal.
6. A raking receiver in a wireless network, as recited in claim 5,
wherein the main path signal processing circuit further comprises:
a main path analog-to-digital converter for converting the main
integrated signal to a digital bit value.
7. A raking receiver in a wireless network, as recited in claim 5,
wherein the main path signal processing circuit further comprises:
a main path amplifier chain for amplifying the main correlated
signal before it is provided to the main path integrating
circuit.
8. A raking receiver in a wireless network, as recited in claim 5,
wherein the main path integrating circuit further comprises: a
first integrator; a second integrator; a third integrator; an input
switch that can select one of the first, second, and third
integrators to connect to the main path mixer; and an output switch
that can select one of the first, second, and third integrators to
connect to a main path analog-to-digital converter, wherein for
every integration period, the input switch selects one of the
first, second, and third integrators as an integrating integrator
to integrate the main correlated signal, the output switch selects
one of the first, second, and third integrators as a holding
integrator to output the main integrated signal, and one of the
first, second, and third integrators operates as a dumping
integrator to dump any information it had previously stored, and
wherein for every integration period, the integrating, holding and
dumping integrators are each different ones of the first, second,
and third integrators.
9. A raking receiver in a wireless network, as recited in claim 1,
wherein phases of each of the first through N.sup.th locally
generated main wavelets can be independently varied by the first
through N.sup.th main path wavelet forming networks,
respectively.
10. A raking receiver in a wireless network, as recited in claim 1,
further comprising: first through M.sup.th error path wavelet
forming networks for producing first through M.sup.th locally
generated error wavelets, respectively; first through M.sup.th
error path weighting mixers for multiplying the first through
M.sup.th locally generated error wavelets by first through M.sup.th
error weighting values, respectively, to produce first through
M.sup.th error weighted wavelets; an error path summmer for adding
together the first through M.sup.th error weighted wavelets to
produce a weighted error correlation input signal; an error path
mixer for multiplying the incoming signal with the weighted error
correlation input signal to produce an error correlated signal; and
an error path signal processing circuit for receiving the error
correlated signal and producing a digital error-tracking signal,
wherein M is an integer greater than 1.
11. A raking receiver in a wireless network, as recited in claim
10, wherein the error path signal processing circuit further
comprises: an error path filtering circuit for filtering the error
correlated signal to produce an error filtered signal.
12. A raking receiver in a wireless network, as recited in claim
11, wherein the error path signal processing circuit further
comprises: an error path analog-to-digital converter for converting
the error filtered signal to a digital bit value.
13. A raking receiver in a wireless network, as recited in claim
11, wherein the error path signal processing circuit further
comprises: an error path amplifier chain for amplifying the error
correlated signal before it is provided to the error path filtering
circuit.
14. A raking receiver in a wireless network, as recited in claim
10, wherein the error path signal processing circuit further
comprises: an error path integrating circuit for integrating the
error correlated signal to produce an error integrated signal.
15. A raking receiver in a wireless network, as recited in claim
14, wherein the error path signal processing circuit further
comprises: an error path analog-to-digital converter for converting
the error integrated signal to the digital error-tracking
signal.
16. A raking receiver in a wireless network, as recited in claim
14, wherein the error path signal processing circuit further
comprises: an error path amplifier chain for amplifying the error
correlated signal before it is provided to the error path
integrating circuit.
17. A raking receiver in a wireless network, as recited in claim
15, wherein the error path integrating circuit further comprises: a
first integrator; a second integrator; a third integrator; an input
switch that can select one of the first, second, and third
integrators to connect to the error path amplifier chain; and an
output switch that can select one of the first, second, and third
integrators to connect to the error path analog-to-digital
converter, wherein for every integration period, the input switch
selects one of the first, second, and third integrators as a sample
integrator to sample the error amplified signal, the output switch
selects one of the first, second, and third integrators as a hold
integrator to output the error integrated signal, and one of the
first, second, and third integrators operates as a dump integrator
to dump any information it had previously stored, and wherein for
every integration period, the sample, hold and dump integrators are
each different ones of the first, second, and third
integrators.
18. A raking receiver in a wireless network, as recited in claim
10, wherein phases of each of the first through M.sup.th error
wavelets can be independently varied by the first through M.sup.th
error path wavelet forming networks, respectively.
19. A raking receiver in a wireless network, as recited in claim
10, wherein M is equal to N.
20. A raking receiver in a wireless network, comprising: an antenna
for receiving an incoming signal; a wavelet forming circuit for
producing first through N.sup.th weighted main wavelets and first
through N.sup.th weighted error wavelets; a main path summer for
adding together the first through N.sup.th weighted main wavelets
to produce a main correlation input signal; an error path summer
for adding together the first through N.sup.th weighted error
wavelets to produce an error correlation input signal; a main path
mixer for multiplying the incoming signal with the main correlation
input signal to produce a main correlated signal; an error path
mixer for multiplying the incoming signal with the error
correlation input signal to produce an error correlated signal; a
main path signal processing circuit for receiving the main
correlated signal and producing a digits bit value; and an error
path signal processing circuit for receiving the error correlated
signal and producing a digital error-tracking signal, wherein N is
an integer greater than 1.
21. A raking receiver in a wireless network, as recited in claim
20, wherein the main path signal processing circuit further
comprises: a main path amplifier chain for amplifying the main
correlated signal to produce a main amplified signal; a main path
integrator circuit fOr integrating the main amplified signal to
produce a main integrated signal; and a main path analog-to-digital
converter for converting the main integrated signal to the digital
bit value.
22. A raking receiver in a wireless network, as recited in claim
21, wherein the main path integrator circuit farther comprises: a
first integrator; a second integrator; a third integrator; an input
switch that can select one of the first, second, and third
integrators to connect to the main path amplifier chain; and an
output switch that can select one of the first, second, and third
integrators to connect to the main path analog-to-digital
converter, wherein for every integration period, the input switch
selects one of the first, second, and third integrators as a sample
integrator to sample the main amplified signal, the output switch
selects one of the first, second, and third integrators as a hold
integrator to output the main integrated signal, and one of the
first, second, and third integrators operates as a dump integrator
to dump any information it had previously stored, and wherein for
every integration period, the sample, hold and dump integrators are
each different ones of the first, second, and third
integrators.
23. A raking receiver in a wireless network, as recited in claim
20, wherein the error path signal processing circuit further
comprises: an error path amplifier chain for amplifying the error
correlated signal to produce an error amplified signal; an error
path integrator circuit for integrating the error amplified signal
to produce an error integrated signal; and an error path
analog-to-digital converter for converting the error integrated
signal to the digital error-tracking signal.
24. A raking receiver in a wireless network, as recited in claim
23, wherein the error path integrator circuit further comprises: a
first integrator; a second integrator; a third integrator; an input
switch that can select one of the first, second, and third
integrators to connect to the error path amplifier chain; and an
output switch that can select one of the first, second, and third
integrators to connect to the error path analog-to-digital
converter, wherein for every integration period, the input switch
selects one of the first, second, and third integrators as a sample
integrator to sample the error amplified signal, the output switch
selects one of the first, second, and third integrators as a hold
integrator to output the error integrated signal, and one of the
first, second, and third integrators operates as a dump integrator
to dump any information it had previously stored, and wherein for
every integration period, the sample, hold and dump integrators are
each different ones of the first, second, and third
integrators.
25. A raking receiver in a wireless network, as recited in claim
20, wherein phases of each of the first through N.sup.th weighted
error wavelets can be independently varied by the wavelet forming
circuit.
26. A method of raking in a wireless network, comprising: receiving
a wireless signal; generating first through N.sup.th locally
generated main wavelets; multiplying the first through N.sup.th
locally generated main wavelets by first through N.sup.th main
weighting values, respectively, to produce first through N.sup.th
main weighted wavelets, respectively, adding together the first
through N.sup.th main weighted wavelets to produce a main
correlation input signal; and correlating the main correlation
input signal with the received signal to produce a bit value,
wherein N is an integer greater than 1.
27. A method of raking in a wireless network, as recited in claim
26, further comprising: generating first through M.sup.th locally
generated error wavelets; multiplying the first through M.sup.th
locally generated error wavelets by first through M.sup.th error
weighting values, respectively, to produce first through M.sup.th
error weighted wavelets; adding together the first through M.sup.th
error weighted wavelets to produce a error correlation input
signal; and correlating the error correlation input signal with the
received signal to produce an error-tracking signal, wherein M is
an integer greater than 1.
28. A method of raking in a wireless network, as recited in claim
27, wherein M is equal to N.
Description
BACKGROUND OF THE INVENTION
Most radios operate in multipath environments. In such multipath
environments, more than one transmission path exists between the
transmitter and receiver. Narrowband radios suffer in multipath
environments due to frequency selective fading, which is caused by
the different time delays on the various paths and the destructive
combining of the signal from all the paths. Narrowband radios can
employ rake receiver structures to combine signals from the
multiple paths, but this is a difficult and expensive process since
narrowband systems lack the time-domain resolution to easily
resolve the multipath terms. Rake is a term used to describe the
coherent combining of energy from a plurality of multi-path induced
replicas of the desired signal.
By definition, however, ultrawide bandwidth (UWB) systems have high
time-domain resolution, and thus can resolve multipath signals.
High chipping rate UWB systems have the advantage of operating in
quasi-stationary multipath environments where the multipath is
changing much slower than the code duration.
A raking receiver is thus used when multiple paths exist between
two radios. FIG. 1 is a block diagram of a wireless system having
two radios in which there are multiple transmission paths between
the two radios.
As shown in FIG. 1, the wireless system 100 includes first and
second radios 110 and 120, having first and second antennas 115,
125, respectively. There is a direct line of sight path 140 between
the two radios 110 and 120, but there are also indirect paths 150
and 155 caused by bouncing signals off of other objects 130, 135 in
the area around the two radios 110 and 120.
As a result, if the first radio 110 sends a wavelet out of the
first antenna 115, the second antenna 125 will receive a plurality
of wavelets having an arbitrary spacing that correspond to that
signal as it passes along one direct path signal 140 and multiple
different reflected paths 150 and 155. And although FIG. 1 shows
only two reflected signals 150 and 155 bouncing off of two objects
130 and 135, there can be many more reflections off of multiple
other objects. In rooms you can have hundreds, even thousands, of
reflections with all kinds of different reflected path lengths.
Furthermore, depending on the properties of each object 130, 135,
the strongest signal received at the second antenna 125 may be a
reflected signal 150, 155 rather than the direct signal 140. One
reason for this is that there could be something collecting energy
at one of the objects 130, 135 and focusing it towards the
receiving antenna 125. Another reason that a reflected signal may
stronger than a direct signal is that there could be multiple
objects that cause reflections having the same reflected path
length. For example, if path 150 has a length L1, path 155 has a
length L2, and L2=L2, the path lengths will be exactly matched. Ma
result of this, one wavelet will travel from the first antenna 115
along path 150 to the second antenna 125, and another wavelet will
travel from the first antenna 115 along path 155 to the second
antenna. But since the path lengths are the same, both wavelets
will arrive at the second antenna 125 at the same time and they
would add their strengths together. Therefore it's not necessary
that the shortest path signal be the strongest one received at the
receiver.
FIGS. 2A-2C are graphs showing examples of the strengths of
received signals in a multipath environment. In particular, FIGS.
2A-2C show the strengths of signals received at the second antenna
125 when a single wavelet is output from the first antenna 115 and
travels only along the three paths 140, 150, and 155 of FIG. 1.
As shown in FIG. 2A, three wavelets 205, 210, and 215 arrive when
the paths 140, 150, and 155 are of different length and the signal
strengths are about the same size. FIG. 2B shows three wavelets
coming in 220, 225, and 230 where the paths 140, 150, and 155 are
of different length and the signal strength of one path is much
larger that the other paths. As a result, one of the wavelets 230
is larger than the other two. FIG. 2C shows only two wavelets 240
and 245 being received because the wavelets from the two reflection
paths 150 and 155 have the same path length (i.e.,
L.sub.1=L.sub.2). As a result, the two reflected wavelets add their
strength and so the second wavelet 245 in this instance is larger
than the first wavelet 240 from the direct path 140.
SUMMARY OF THE INVENTION
Consistent with the title of this section, only a brief description
of selected features of the present invention is now presented. A
more complete description of the present invention is the subject
of this entire document.
An object of the present invention is to maximize the amount of
rake through multiple phases of an incoming signal, while
minimizing the circuit complexity and power consumption.
Another object of the present invention is to provide signals for a
signal correlation path and an error/tracking path within a rake
channel with a minimum of circuitry and power consumption.
These and other objects are accomplished by way of a raking
receiver in a wireless network. The raking receiver comprises: an
antenna for receiving an incoming signal; first through N.sup.th
main path wavelet forming networks for producing first through
N.sup.th locally generated main wavelets; first through N.sup.th
main path weighting mixers for multiplying the first through
N.sup.th locally generated main wavelets by first through N.sup.th
main weighting values, respectively, to produce first through
N.sup.th main weighted wavelets; a main path summer for adding
together the first through N.sup.th main weighted wavelets to
produce a weighted main correlation input signal; a main path mixer
for multiplying the incoming signal with the weighted main
correlation input signal to produce a main correlated signal; and a
main path signal processing circuit for receiving the main
correlating circuit and producing a digital bit value, wherein N is
an integer greater than 1.
The main path signal processing circuit may further comprise: a
main path filtering circuit for filtering the main correlated
signal to produce a main filtered signal; a main path
analog-to-digital converter for converting the main filtered signal
to a digital bit value; and a main path amplifier chain for
amplifying the main correlated signal before it is provided to the
main path filtering circuit.
The main path signal processing circuit may further comprise: a
main path integrating circuit for integrating the main correlated
signal to produce a main integrated signal; a main path
analog-to-digital converter for converting the main integrated
signal to a digital bit value; and a main path amplifier chain for
amplifying the main correlated signal before it is provided to the
main path integrating circuit.
The main path integrator may further comprise: a first integrator;
a second integrator; and a third integrator; an input switch that
can select one of the first, second, and third integrators to
connect to the main path mixer; and an output switch that can
select one of the first, second, and third integrators to connect
to the main path analog-to-digital converter. In this case, for
every integration period, the input switch preferably selects one
of the first, second, and third integrators as an integrating
integrator to integrate the main correlated signal, the output
switch preferably selects one of the first, second, and third
integrators as a holding integrator to output the main integrated
signal, and one of the first, second, and third integrators
preferably operates as a dumping integrator to dump any information
it had previously stored. And for every integration period, the
integrating, holding and dumping integrators are preferably each
different ones of the first, second, and third integrators.
The phases of each of the first through N.sup.th main wavelets can
preferably be independently varied by the first through N.sup.th
main path wavelet forming networks, respectively.
The raking receiver may further comprise: first through M.sup.th
error path wavelet forming networks for producing first through
M.sup.th locally generated error wavelets; first through M.sup.th
error path weighting mixers for multiplying the first through
M.sup.th locally generated error wavelets by first through M.sup.th
error weighting values, respectively, to produce first through
M.sup.th error weighted wavelets; an error path summer for adding
together the first through M.sup.th error weighted wavelets to
produce a weighted error correlation input signal; an error path
mixer for multiplying the incoming signal with the weighted error
correlation input signal to produce an error correlated signal; an
error path signal processing circuit for receiving the error
correlating circuit and producing a digital error-tracking signal,
wherein M is an integer greater than 1.
The error path signal processing circuit may further comprise: an
error path filtering circuit for filtering the error correlated
signal to produce an error filtered signal; an error path
analog-to-digital converter for converting the error filtered
signal to a digital bit value; and an error path amplifier chain
for amplifying the error correlated signal before it is provided to
the error path filtering circuit.
The error path signal processing circuit may further comprise: an
error path integrating circuit for integrating the error correlated
signal to produce an error integrated signal; an error path
analog-to-digital converter for converting the error integrated
signal to a digital bit value; and an error path amplifier chain
for amplifying the error correlated signal before it is provided to
the error path integrating circuit.
The error path integrator circuit may further comprise: a first
integrator; a second integrator; a third integrator; an input
switch that can select one of the first, second, and third
integrators to connect to the error path amplifier chain; and an
output switch that can select one of the first, second, and third
integrators to connect to the error path analog-to-digital
converter. For every integration period, the input switch
preferably selects one of the first, second, and third integrators
as a sample integrator to sample the error amplified signal, the
output switch preferably selects one of the first, second, and
third integrators as a hold integrator to output the error
integrated signal, and one of the first, second, and third
integrators preferably operates as a dump integrator to dump any
information it had previously stored. And for every integration
period, the sample, hold and dump integrators are preferably each
different ones of the first, second, and third integrators.
The phases of each of the first through M.sup.th error wavelets can
preferably be independently varied by the first through M.sup.th
error path wavelet forming networks, respectively. Also, M may be
equal to N.
An alternate raking receiver in a wireless network is also
provided. This raking receiver comprises: an antenna for receiving
an incoming signal; A wavelet forming circuit for producing first
through Nth weighted main wavelets and first through Nth weighted
error wavelets; A main path summer for adding together the first
through Nth weighted main wavelets to produce a main correlation
input signal; an error path summer for adding together the first
through Nth weighted error wavelets to produce an error correlation
input signal; a main path mixer for multiplying the incoming signal
with the main correlation input signal to produce a main correlated
signal; an error path mixer for multiplying the incoming signal
with the error correlation input signal to produce an error
correlated signal; a main path signal processing circuit for
receiving the maim correlated signal and producing a digital bit
value; an error path signal processing circuit for receiving the
error correlating circuit and producing a digital error-tracking
signal, wherein N is an integer greater than 1.
The main path signal processing circuit may further comprise: a
main path amplifier chain for amplifying the main correlated signal
to produce a main amplified signal; a main path integrator circuit
for integrating the main amplified signal to produce a main
integrated signal; and a main path analog-to-digital converter for
converting the main integrated signal to the digital bit value.
The main path integrator circuit may further comprise: a first
integrator; a second integrator; a third integrator; an input
switch that can select one of the first, second, and third
integrators to connect to the main path amplifier chain; and an
output switch that can select one of the first, second, and third
integrators to connect to the main path analog-to-digital
converter. For every integration period, the input switch
preferably selects one of the first, second, and third integrators
as a sample integrator to sample the main amplified signal, the
output switch preferably selects one of the first, second, and
third integrators as a hold integrator to output the main
integrated signal, and one of the first, second, and third
integrators preferably operates as a dump integrator to dump any
information it had previously stored. And for every integration
period, the sample, hold and dump integrators are preferably each
different ones of the first, second, and third integrators.
The error path signal processing circuit may further comprise: an
error path amplifier chain for amplifying the error correlated
signal to produce an error amplified signal; an error path
integrator circuit for integrating the error amplified signal to
produce an error integrated signal; and an error path
analog-to-digital converter for converting the error integrated
signal to the digital error-tracking signal.
The error path integrator circuit may further comprise: a first
integrator; a second integrator; a third integrator; an input
switch that can select one of the first, second, and third
integrators to connect to the error path amplifier chain; and an
output switch that can select one of the first, second, and third
integrators to connect to the error path analog-to-digital
converter. For every integration period, the input switch
preferably selects one of the first, second, and third integrators
as a sample integrator to sample the error amplified signal, the
output switch preferably selects one of the first, second, and
third integrators as a hold integrator to output the error
integrated signal, and one of the first, second, and third
integrators preferably operates as a dump integrator to dump any
information it had previously stored. And for every integration
period, the sample, hold and dump integrators are preferably each
different ones of the first, second, and third integrators.
The phases of each of the first through M.sup.th error wavelets can
preferably be independently varied by the first through M.sup.th
error path wavelet forming networks, respectively.
A method of raking in a wireless network is provided. The method
comprises: receiving a wireless signal; generating first through
N.sup.th locally generated main wavelets; multiplying the first
through N.sup.th locally generated main wavelets by first through
N.sup.th main weighting values, respectively, to produce first
through N.sup.th main weighted wavelets; adding together the first
through N.sup.th main weighted wavelets to produce a main
correlation input signal; and correlating the main correlation
input signal with the received signal to produce a bit value,
wherein N is an integer greater than 1.
The method may further comprise: generating first through M.sup.th
locally generated error wavelets; multiplying the first through
M.sup.th locally generated error wavelets by first through M.sup.th
error weighting values, respectively, to produce first through
M.sup.th error weighted wavelets; adding together the first through
M.sup.th error weighted wavelets to produce a error correlation
input signal; and correlating the error correlation input signal
with the received signal to produce an error-tracking signal,
wherein M is an integer greater than 1. M may be equal to N.
BRIEF DESCRIPTION OF THE DRAWINGS
A more complete appreciation of the invention and its many
attendant advantages will be readily obtained as it becomes better
understood with reference to the following detailed description
when considered in connection with the accompanying drawings, in
which:
FIG. 1 is a block diagram of a wireless system having two radios in
which there are multiple transmission paths between the two
radios;
FIGS. 2A-2C are graphs showing examples of the strengths of
received signals in a multipath environment;
FIG. 3 is a block diagram of a wireless system including a rake
receiver according to a preferred embodiment of the present
invention;
FIG. 4A is a diagram of a portion of a network including a rake
radio;
FIG. 4B is a diagram of an integrator circuit from the network of
FIG. 4B according to a preferred embodiment of the present
invention;
FIG. 5 is signal graph of signals coming into the first channel of
the receiver of FIG. 4A;
FIG. 6 is signal graph of signals being correlated by the main path
of the first radio in FIG. 4A;
FIG. 7 is a block diagram of a wireless network including a rake
receiver according to a preferred embodiment of the present
invention;
FIG. 8 is a graph of a wireless wavelet as it is reflected off of
two surfaces in sequence;
FIG. 9 is a block diagram showing a rake receiver according to a
preferred embodiment of the present invention;
FIGS. 10A and 10B are block diagrams of combined wave-generator and
weighting circuits according to preferred embodiments of the
present invention;
FIG. 11 is a top-level block diagram of a raking system according
to a preferred embodiment of the present invention;
FIG. 12 is a top-level block diagram of a raking system according
to another preferred embodiment of the present invention;
FIG. 13 is a block diagram of a multiple rake receiver according to
a preferred embodiment of the present invention;
FIG. 14 is a block diagram of a controller and wave-generator
system for use in the multiple rake receiver of FIG. 13; and
FIG. 15 is a flow chart of a raking method according to a preferred
embodiment of the present invention.
DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS
Preferred embodiments of the present invention will now be
described with reference to the drawings. Throughout the several
views, like reference numerals designate identical or corresponding
parts.
In operation a rake processor or rake receiver effectively turns
two or more radios into one super high performance radio. Using the
example of FIGS. 1 and 2B, the rake receiver could have three
effective radios--one tuned to the first incoming wavelet 220 of
FIG. 2A, one tuned to the second incoming wavelet 225, and one
tuned to the third incoming wavelet 230--and coherently combine the
outputs of all three radios so as to effectively receive a much
better signal.
FIG. 3 is a block diagram of a wireless system including a rake
receiver according to a preferred embodiment of the present
invention. As shown in FIG. 3, the raking receiver 320 includes
first, second, and third radio receivers 360, 370, and 380, each
sharing the same antennae 325. Each of the receivers 360, 370, 380
can be called a channel, a finger, or an arm in the rake
correlator. In this preferred embodiment the antennae 325 is
attached to the first, second, and third radio receivers 360, 370,
and 380 in the raking receiver 320, each of which radio receiver is
tuned to a specific wavelet 220, 225, and 230, respectively from
FIG. 2B.
The first, second and third radio receivers 350, 360, 370 output
first, second, and third output signals, respectively, which are
then weighted by first, second, and third weight values W.sub.1,
W.sub.2, or W.sub.3, respectively. The weighted outputs are then
added together in a summer 390, which outputs a final bit
value.
The first, second, and third weights W.sub.1, W.sub.2, and W.sub.3
used in the rake receiver 320 if FIG. 3 are used to account for the
variance and phase of the signal. If a huge signal (i.e. with low
variance or relatively low noise) is received at one receiver
(e.g., the incoming wavelet 230 at the third radio receiver 380),
that signal will be weighted very high (e.g., W.sub.3 will be
comparatively high). This higher weight indicates that the signal
received at that particular radio receiver is more likely to be the
correct signal, and the least likely to be corrupted by noise or
interference.
The radio receiver with the highest weight will indicate the value
of a received bit with the lowest error rate. The other radios 360
and 370 (which receive noisier signals and therefore have smaller
weights W.sub.1 and W.sub.2 as compared to W.sub.3) provide a bit
value that has a higher likelihood of being in error. But if the
system combines the three weighted signals, the combined result can
be more accurate than just the output of any one of the radios by
itself--even the radio with the lowest error.
In order to achieve this increased accuracy, the outputs of the
radios 360, 370, and 380 are preferably weighted in inverse
proportion to their variance. In other words, something of higher
variance (e.g., wavelets 220 and 225) would be weighted less, while
something of lower variance (e.g., wavelet 230) would be weighted
more.
Although this preferred embodiment discloses three radio receivers,
alternate embodiments could expand this to N radio receivers, where
N is an integer greater than one. In this case each of the N
receiver outputs will be weighted by a respective first through
N.sup.th weight value W.sub.1-W.sub.N.
One problem with this design, however, is that it requires that the
receiver to have N entire radios. For example, if a rake receiver
of the sort shown in FIG. 3 were designed to rake sixteen terms, it
would have to use sixteen separate radios, figure out the variance
of each of the sixteen incoming signals, establish sixteen weights,
and sum all sixteen weighted signals to achieve a final result.
Larger raking numbers would require increased complexity.
Computing Variance and Weights
With respect to determining variance and weights, the receiver will
preferably include a separate variance determining circuit that
would determine the variance and the weights dynamically or at
least partially dynamically. In other words, the circuit will
preferably determine the variance for each incoming signal as it is
received.
Some methods of determining variance are shown in: application Ser.
No. 09/685,197 filed Oct. 10, 2000, entitled MODE CONTROLLER FOR
SIGNAL ACQUISITION AND TRACKING IN AN ULTRA WIDEBAND COMMUNICATION
SYSTEM; application Ser. No. 09/685,195 filed Oct. 10, 2000,
entitled ULTRA WIDE BANDWIDTH SYSTEM AND METHOD FOR FAST
SYNCHRONIZATION; application Ser. No. 09/684,401 filed Oct. 10,
2000, entitled ULTRA WIDE BANDWIDTH SYSTEM AND METHOD FOR FAST
SYNCHRONIZATION USING SUB CODE SPINS; and application Ser. No.
10/214,183 filed Aug. 8, 2002, entitled MODE CONTROLLER FOR SIGNAL
ACQUISITION AND TRACKING IN AN ULTRA WIDEBAND COMMUNICATION SYSTEM,
all of which are incorporated by reference into the current
application.
In the alternative, a variance determining circuit in the receiver
could examine an incoming signal for some period of time, compute a
variance, set a proper weight value, and then go to sleep for
awhile. The weights can thus be set periodically in some
embodiments because the variances for the various signals are
unlikely to change very quickly.
One example of a partially dynamic weighting process is one that is
performed on a packet-by-packet basis. The variance estimation and
weight determining circuit operates during the preamble of the
packet, and weights determined are used for the remainder of the
packet. In other words, the variance and weight determination
circuit could wake up when the preamble starts, listen to the
incoming signal, calculate the variances, set the appropriate
weights, and then leave the weight values unchanged for the entire
data packet. For the duration of that packet, data will come in at
each separate radio receiver (i.e., at each channel of the rake
receiver) and will be weighted based on the variances determined at
the beginning of the packet. Since packets are generally sent
frequently (every 10's to 100's of a millionth of one second), the
multipath appears stationary over the duration of the packet, so
the ideal weights remain constant throughout the packet.
In alternate embodiments the device may update the weights less
often, e.g., every P number of packets (where P is an integer
greater than 1). After every P packets the device will re-estimate
the variances and determine the appropriate weights. In between
these P packets, the device simply freezes the weights at their
current values.
The precise implementation of the variance determining circuit can
depend upon how quickly the channel is expected to change. For
example, if the first and second antennas 115 and 125 from FIG. 1
are on rooftops and the interfering structures 130, 135 are
buildings, then the parameters of the channel would change very
slowly, weeks to years (e.g., it might change if a building were
torn down or another one erected). But if the devices 110 and 120
are handheld devices whose users are walking around a room, then
the interfering structures 130, 135 may change quickly relative to
the moving radios 110 and 120. Moreover, the interfering structures
130, 135 could also be moving (e.g., they could be other people.)
causing the channel parameters to alter frequently. In this case it
may be necessary to update the variances and weights every
hundredth of a second.
Raking Receiver
FIG. 4A is a diagram of a portion of a network including a rake
radio. As shown in FIG. 4A, the network includes a first device 110
having a first antenna 115 and a second device 420 having a second
antenna 425. The second device 420 includes first and second radio
receivers (also called channels, fingers, or arms) 430a and 430b,
and a summer 435. The first radio receiver 430a includes a first
main mixer 440a, a first error mixer 445a, a first main wavelet
forming network (wave-generator) 450a, a first error wave-generator
455a, a first main amplifier chain 460a, a first error amplifier
chain 465a, a first main integrate-hold-dump (IHD) integrator 470a,
a first error IHD integrator 475a, a first main analog-to-digital
converter (ADC) 480a, a first error ADC 485a, a first main
weighting mixer 490a, and a first error weighting mixer 495a. The
second radio receiver 430b includes a second main mixer 440b, a
second error mixer 445b, a second main wave-generator 450b, a
second error wave-generator 455b, a second main amplifier chain
460b, a second error amplifier chain 465b, a second main IHD
integrator 470b, a second error IHD integrator 475b, a second main
ADC 480b, second error ADC 485b, a second main weighting mixer
490b, and a second error weighting mixer 495b.
Some examples of designs for phase code demodulators 451a, 451b,
456a, 456b may be found in: application Ser. No. 09/209,460 filed
Dec. 11, 1998, entitled ULTRAWIDE BANDWIDTH SPREAD-SPECTRUM
COMMUNICATIONS SYSTEM; application Ser. No. 09/972,966 filed Oct.
10, 2001, entitled ULTRA WIDE BANDWIDTH NOISE CANCELLATION
MECHANISM AND METHOD; application Ser. No. 09/684,400 filed Oct.
10, 2000, entitled ULTRAWIDEBAND COMMUNICATION SYSTEM, METHOD, AND
DEVICE WITH LOW NOISE PULSE FORMATION; application Ser. No.
09/684,782 filed Oct. 10, 2000, entitled ULTRA WIDEBAND
COMMUNICATION SYSTEM, METHOD, AND DEVICE WITH LOW NOISE RECEPTION;
and application Ser. No. 09/685,200 filed Oct. 10, 2000, entitled
LEAKAGE NULLING RECEIVER CORRELATOR STRUCTURE AND METHOD FOR ULTRA
WIDE BANDWIDTH COMMUNICATION, all of which are incorporated by
reference into the current application.
The first device 110 in FIG. 4A sends a signal to the second device
420 along a plurality of one or more signal paths, depicted as 140,
150, and 155. The receiving antenna 425 receives the incoming
signal from all of the signal paths 140, 150, 155, splits it into
two copies, and sends the incoming signal to each of radio
receivers 430a and 430b, to both the main and error paths in
each.
The raking receiver includes 2 or more channels. As shown in the
preferred embodiment of FIG. 4A, the raking receiver 420 includes
two channels: a first channel processed by the first radio receiver
430a, and a second channel processed by the second radio receiver
430b. By way of example, the following discussion will examine the
operation of the first channel (run by the first radio 430a). The
operation of the second channel (run by the second radio 430b) will
be analogous.
In the first channel, the incoming signal comes in at the second
antenna 425 and passes through the first dual phase code
demodulator block 452a, into the first main phase code demodulator
block 451a and the first error phase code demodulator block 456a.
Here the incoming signal passes to mixers 440a and 445a. The first
main wave-generator 450a, connected to the first main mixer 440, is
clocked at or near the chipping rate of the incoming signal. While
the clocking rate is near the chipping rate of the incoming signal,
the wave-generator 450a scans through possible phases so that the
first main path will find all of the incoming wavelets (e.g., as
shown in FIGS. 2A-2C).
Having found the clock phase for each of the incoming wavelets, the
first main wave-generator 450a and first error wave-generator 455a
will clock at the phase that aligns the largest of the incoming
signal wavelets and the wavelet from wave-generator 450 at mixer
440a so that the first channel 430a receives the energy from the
largest wavelet in the incoming signal (e.g., signal 230 from FIG.
2B). First error wave-generator 455a generates a quadrature signal
that is aligned with the incoming signal at mixer 445a so as to
produce a null signal when the phase alignment is correct, and a
plus or minus signal if the phase is early or late. The first radio
430a will then use the signal traveling on the error path (elements
445a, 455a, 465a, 475a, 485a, 495a) to track the phase and maintain
the alignment of the incoming signal with the wavelet from
wave-generator 450a at mixer 440a, and with the wavelet from
wave-generator 455a at mixer 445a.
In some preferred embodiments, the various wave generators 450a,
450b, 455a, 455b can be pulse form generators, sine wave
generators, or the like.
In a preferred embodiment, during a scanning process to find the
phase for the plurality of multipath terms, each phase code
demodulator 451a, 451b, 456a, 456b, can be run at phase offsets
relative to each other in order to accelerate the search for the
largest multipath terms.
FIG. 5 is signal graph of signals coming into the first channel of
the raking receiver 420 of FIG. 4A. FIG. 6 is signal graph of
signals being correlated by the main path of the first radio in
FIG. 4A. As shown in FIG. 5, wavelets 220, 225, and 230 represent
incoming wavelets, while wavelets 550a and 550b represent a wavelet
generated by the first main wave-generator 450a at various phases
(i.e., as the first main wave-generator 450a varies its phase).
When the wavelet initially generated by the main wave-generator
550a matches the phase of a wavelet received at the antenna (e.g.,
wavelet 550b matches phase with wavelet 230), the two wavelets
550b, and 230 will correlate at the first main mixer 440a as shown
in FIG. 6.
During correlation of the incoming wavelet 230 and the locally
generated wavelet 550b, the positive portion in the incoming
wavelet 230 will combine with the positive portion in the wavelet
550b generated by the first main wave-generator 450a to make a
positive signal 610. Likewise, the negative portion in the incoming
wavelet 230 will combine with the negative portion in the wavelet
550b generated by the first wave-generator 450 to make a positive
signal 620. As a result the first main mixer 440a will output a
double-humped wavelet (wavelet 610 plus wavelet 620). Since this
double-humped wavelet (wavelet 610 plus wavelet 620) is either all
positive or all negative, when passed into integrator (like 470a in
FIG. 4B) a DC value is produced that is proportional to the signal
energy in the multipath term being received, and the polarity is
consistent with the phase of the transmitted information (e.g. plus
or minus for a digital one or zero). If the transmitter (110 of
FIG. 4B) is transmitting a code sequence of phase modulated
wavelets, where the code duration is N wavelets (N=1 or more), then
the first radio can take the output of the integration of N
wavelets and use it as the magnitude and phase (or polarity if it
is bi-phase) of the signal.
This operation is performed along the main path in the first radio
430a. The incoming signal is passed to the first main mixer 440a,
which correlates the incoming signal with the local signal
generated by the first main wave-generator 450a. The result of this
is sent through the fist amplifier chain 460a, through the first
main IHD integrator, through the first main ADC 480a, and finally
through the first main weighting mixer 490a (where it will be
weighted by a first weighting value W.sub.1). The result is then
sent to the summer 345 where it is summed with the weighted result
from the second radio 430b to produce a final value.
In some embodiments of an ultrawide bandwidth (UWB) radio, each
transmitted bit is represented by a single wavelet. In this case,
when the radio receives a single wavelet, it has received a
transmitted bit. A UWB system generally has a very, very high
spatial resolution because the wavelet is so short in time.
Furthermore, although FIGS. 2 and 5 show each of the signals being
very clearly identifiable as separate wavelets, it is possible that
wavelets from multiple paths could overlap one another. In this
case the receiver would receive a different waveform that would be
a composite of the wavelets. But when that happens, the resulting
waveform will still correlate with the wavelet generated by the
main wave-generator 450a, and therefore produce a useable output
signal that corresponds to the phase of the information on the
incoming signal.
As a result of this, the first and second channels (represented by
the first and second radios 430a and 430b) will separately
correlate with different multipath signals. For example, the first
channel (i.e., the first radio 430a) might lock onto the third
wavelet 230 of FIG. 2B, while the second channel (i.e., the second
radio 430b) might lock onto the second wavelet 225 in FIG. 2B. Each
radio 430a, 430b will then correlate and determine some value for
the correlation (as shown in FIG. 6).
FIG. 4B is a block diagram of an integrator of FIG. 4A according to
a preferred embodiment of the present invention. As shown in FIG.
4B, a preferred IHD integrator 470a, 470b, 475a, 475b comprises
first, second, and third integrators 471, 472, 473, an input switch
476 that choose one of the first, second, and third integrators
471, 472, 473, and an output switch 477 that choose one of the
first, second, and third integrators 471, 472, 473.
In the embodiment shown in FIG. 4A, each path of each of the
channels has to have a phase-code demodulator (comprised of a wave
generator, a mixer, an amplifier chain), an integrator, a one or
more bit ADC, and a mechanism to weight the result. In the
preferred embodiment (as shown in FIG. 4B), the integrator 470a,
470b, 475a, 475b may be comprised of an integrate-hold-dump (IHD)
integrator that time shares three individual integrators 471, 472,
473 where one is integrating over the transmitted phase-code
interval, while another is dumping (i.e. discharging to zero),
while the remaining one is holding the result of the previous
integration which is passed to the ADC to digitally capture the
signal value. After each integration period, the functions of the
three integrators rotate so the integrator that was holding, dumps;
the integrator that was dumping, integrates; and the integrator
that was integrating, holds.
The weighting can be done digitally, as shown in FIG. 4A with
digital multiplier 490a, 490b, 495a, 495b. It can also be done
within the analog processing with a variable gain and
phase-reversal function within the amplifier chain 460a, 460b,
465a, 465b. Similarly, it could be done with a multiplying ADC. It
could also be split where the gain is applied in one or more of the
analog processing blocks, while the sign was applied in the digital
blocks. To support either an error path, or a quadrature path to
support QPSK, or M-ary PSK or M-ary QAM modulated wavelets, or FSK,
each channel has to have two paths (i.e. two mixers, two amplifier
chains, two integrators (IHD or otherwise), two ADCs, and two
weighting mechanisms).
If an embodiment had sixteen channels (i.e. letters "a" through
"p", each of these channel structures 430a-430p would have to have
the above components be repeated, which would amount to thirty-two
mixers, thirty-two amplifier chains, thirty-two integrators
(ninety-six integrators if using IHD integrators), thirty-two ADCs,
and thirty-two weighting mechanisms. Such a collection of
components would use a lot of power and would take up a lot of
space on a chip.
It would be desirable to minimize the number of components used for
a rake receiver, however. This will keep the integrated circuit die
size small, which will in turn reduce costs since die size is
proportional to costs.
Raking Process
In the system disclosed in FIG. 4A, the raking process is performed
as follows. The first device 110 sends a signal including multiple
data bits to the second device 420. This signal travels along
multiple paths (including a direct path 140 and a plurality of
reflected paths such as 150, 155) and is received at the second
device 420.
The received signal (including multiple reflected copies of the
same signal) is split and sent to each channel (i.e., finger or
arm) in the receiving (i.e., second) device 420. The N.sub.c
channels are synchronized to correlate with the N.sub.c largest
wavelet terms in the incoming signal. At this point, the first and
second wave-generators 450a and 450b have different phases, e.g.,
the first wave-generator 450a may be timed at T.sub.1, while the
second wave-generator 450b may be timed at T.sub.2. (This could be
extended up to T.sub.N.sub.C if N.sub.c channels were used.)
And as noted above, in this structure, the first main
wave-generator 450a and the second main wave-generator 450b could
be correlating on wavelets from different transmitted bits. For
example, the first channel could be locking onto one wavelet (e.g.,
the third wavelet 230 from FIG. 2B), while the second channel might
be locking onto a different wavelet (e.g., the second wavelet 225
from FIG. 2B).
Furthermore, the second wavelet 225 might be from a second data
bit, while the third wavelet 230 might actually be from a first
data bit. In other words wavelet 225 might come from path 140 while
wavelet 230 might come from path 150 and the difference in path
length could mean that different bits were received for each. For
example, if the direct path 140 was very short and the first
reflected path 150 was very long, a second wavelet 225 from a
second data bit might arrive on the direct path 140 before a third
wavelet 230 from a first data bit arrived on the first reflected
path 150. In this way, the first bit could arrive at the second
device after the second bit.
Therefore, it is necessary to identify which data bits each channel
has correlated onto. In particular, when the second device 420
works to combine the results from the channels of the raking
receiver, the wave-generators in each channel will need to know
what signal corresponds to what bit.
More generally, the information is encoded in symbols that are
comprised of a family of modulated wavelets which are modulated
according to codes (also called code words). In this case, data
bits are represented not by a single wavelet but by a sequence of
wavelets. Each symbol may contain one or more bits of information.
Each channel in the rake receiver (i.e., the fingers or arms)
aligns to the sequence of wavelets in a symbol rather than just a
single wavelet.
Once the first main wave-generator 450a and the second main
wave-generator 450b are synchronized to the symbol, the resulting
values that they produce can be weighted as discussed above with
respect to FIG. 3. A similar process is performed for each error
path so that each multipath term is tracked separately.
Alternatively, the "error" path can be used as a quadrature channel
to allow M-ary PSK or M-ary QAM to be decoded. Finally, the
weighted results from each channel are added to determine a final
value for the symbol that was sent.
First Alternate Raking Receiver
FIG. 7 is a block diagram of a wireless network including a rake
receiver according to a preferred embodiment of the present
invention. As shown in FIG. 7, the network includes a first device
110 having a first antenna 115 and a second device 720 having a
second antenna 725. The second device includes a single channel 720
that includes two paths, a main path and an error path, where these
could also be called an in-phase path and a quadrature path. The
paths are comprised of a main summer 730, an error summer 735, a
main mixer 740, an error mixer 745, first through N.sup.th main
wavelet forming networks (wave-generators) 750.sub.1-750.sub.N,
first through N.sup.th error wave-generators 755.sub.1-755.sub.N, a
main amplifier chain 760, an error amplifier chain 765, a main
integrate-hold-dump (IHD) integrator 770, an error IHD integrator
775, a main analog-to-digital converter (ADC) 780, an error ADC
785, first through N.sup.th main weighting mixers
790.sub.1-790.sub.N, and first through N.sup.th error weighting
mixers 795.sub.1-795.sub.N. Alternatively, the wave-generators
could be amplitude controlled, eliminating the need for the
weighting mixers. Similarly, the summing block could have
independently weighted inputs, also eliminating the need for the
weighting mixers. Thus, the rake receiver of FIG. 7 includes only a
single radio receiver, but multiple wave-generators
750.sub.1-750.sub.N and 755.sub.1-755.sub.N and weighting mixers
790.sub.1-790.sub.N, 795.sub.1-795.sub.N, or simply multiple
amplitude controlled wave-generators In this preferred embodiment,
N can be any integer value greater than 1, and is preferably
matched to the expected number of multipath terms found where the
radio will be operating, which is generally between 2 and 32.
In operation, the first device 110 in FIG. 7 sends a signal to the
second device 720 along a plurality of signal paths 140, 150, and
155. The receiving antenna 725 receives the incoming signal
containing signals from all of the signal paths 140, 150, 155, and
splits the received incoming signal, sending it to the main and
error paths via the main and error mixers 740 and 745.
Alternatively, these paths could be labeled in-phase and quadrature
paths to support QPSK, or M-ary PSK or M-ary QAM.
The rake receiver of FIG. 7 uses a single radio to perform a rake
function by having multiple wave-generators whose outputs are
weighted (at the main and error weighting mixers
790.sub.1-790.sub.N and 795.sub.1-795.sub.N) and summed (at the
main and error summers 730 and 735) before the results are
amplified, integrated, or converted from analog to digital signals.
And despite the fact that this summing is done before all of these
power consumptive steps, the resulting bit value output from the
main ADC 780 can be nominally the same as the bit value output from
the summer 435 of FIG. 4.
For each path (main and error, or in-phase and quadrature) in FIG.
7, there are N wave-generators 750.sub.1-750.sub.N or
755.sub.1-755.sub.N, and N weights W.sub.1 to W.sub.N applied
through N weighting mixers 790.sub.1-790.sub.N or
795.sub.1-795.sub.N, and a summer 730 and 735 to add the weighted
wavelets together and supply the weighted and summed wave-generator
value to the relevant mixer 740 or 745. The first through N.sup.th
weights W.sub.1 to W.sub.N are preferably the same weights that
would have been put on the end of an N-arm receiver of the sort
shown in FIG. 4. In other words, these are the weights for the
correlation signal.
However, by placing the wave-generators 750.sub.1-750.sub.N and
755.sub.1-755.sub.N, and weights on the wave-generator side of
mixers 740 and 745, the raking receiver can eliminate the circuitry
that would otherwise have to appear after these mixers in each of
the channels. The embodiment of FIG. 7 avoids the need for this
circuitry, leaving just two paths, where each has an amplifier,
integrator, and ADC chain.
Consider a raking receiver as shown in FIG. 4. Each separate radio
430a, 430b in the second (receiving) device 420 requires two
wave-generators, two mixers, two amplifier chains, two IHD
integrators, two ADCs, two weighting circuits, plus the device 420
requires one summer 435 to sum the results from each radio
receiver. If the device 420 had four arms it would require eight
wave-generators, eight weighting mixers, eight signal mixers, eight
amplifier chains, eight integrators (preferably IHD), eight ADCs,
and one summer.
In contrast, if the device 720 had four arms it would require eight
wave-generators and eight weighting mixers, but only two signal
mixers (main and error), two amplifier chains, two integrators
(preferably IHD), two ADCs, and two summers. A raking receiver as
shown in FIG. 7 has a single radio with multiple wave-generators.
Regardless of how many channels (i.e., fingers or arms) the second
(receiving) device 720 uses, it will only have one pair of main and
error paths including just two summers 730, 735, two mixers 740,
745, two sets of amplifier chains 760, 765, two IHD integrators
770, 775, and two ADCs 780, 785. The device 720 only adds two
wave-generators, two signal mixers (main and error) for each added
channel. Extended to N arms, the embodiment of FIG. 7 will always
have the same number of wave-generators and weighting mixers (or
amplitude controlled wave-generators), but only (1/N ).sup.th the
number of main/error mixers, amplifier chains, IHD integrators, and
ADCs.
Thus, while the embodiment of FIG. 7 does not reduce the complexity
of the wave-generator structure (i.e., the agile clock), it does
eliminate a great number of circuit elements, which can
significantly reduce the cost of the device.
In operation, the second (receiving) device 720 performs multiple
processing on the incoming signal, correlating and weighting it
with multiple wave-generators at once. The signals stack on top of
each other and are all mixed with the incoming signal at the same
time. The result of this signal processing is the same as for the
circuit shown in FIG. 4, except that actual processing of the
signals is performed in a different order.
In alternate embodiments the receiving device 720 in FIG. 7 could
process multipath wavelets in the incoming signal individually. In
one particular alternate embodiment, the device 720 could turn off
all of the wave-generators 750.sub.1-750.sub.N, 755.sub.1-755.sub.N
except for one in the main path and one in the error path. Each the
clock on each wave-generator could then be scanned to find the
largest wavelet in the incoming signal. One at a time, each
wave-generator could scan for a correlated multipath term that most
improves the SNR of the final output signal from the ADC.
Of course some of the reflected signals received at the device 720
will be non-inverted, and some will be inverted. This is because
signals can be reversed when they are reflected. An example of this
is shown in FIG. 8. As each wave-generator scans for a correlated
multipath term, the correlation coefficient is either positive or
negative, and is used to set the weight to positive or negative,
which accounts for this multipath reversal.
Reflections
FIG. 8 is a graph of a wireless wavelet as it is reflected off of
two surfaces in sequence. As seen in FIG. 8, a wavelet 810 is
provided, e.g., sent from a first device 710. The wavelet 810 is
then bounced off a first interfering object 850 (such as a first
metal plate). The first reflected wavelet 820 that comes off of the
first interfering object 850 is a derivative of the first wavelet
810 and looks very much like an upside down version of the initial
wavelet 810 (although not identical).
If the second wavelet 820 bounces off of a second interfering
object 860 (such as a second metal plate), it will reflect as a
second reflected wavelet 830 that is the derivative of the second
wavelet 820 and looks very similar to the original wavelet 810
(although not identical).
Wavelets that have bounced an even number of times (0, 2, 4, . . .
) can be called even bounce wavelets, while wavelets that have
bounced an odd number of times (1, 3, 5, . . . ) can be called odd
bounce wavelets. Because of this phenomenon, it is possible to tell
a difference between an odd bounce wavelet and an even bounce
wavelet because the waveforms are inverted on each odd bounce and
are returned to their original polarity on each even bounce.
In the receiving device 720 of FIG. 7, the weights W.sub.1 to
W.sub.N applied to the signal output from the first through
N.sup.th wave-generators can have their polarity adjusted to
account for the polarity of the incoming signals. The normalized
weights W.sub.1 to W.sub.N vary between plus and minus one, with
the values being one polarity for odd bounce wavelets and the other
polarity for even bounce wavelets (e.g., negative for odd bounce
wavelets and positive for even bounce wavelets) This allows the
device 720 to receive signals no matter how many bounces they have.
Thus, the waveforms output from the wave-generators
750.sub.1-750.sub.N, 755.sub.1-755.sub.N will be properly
correlated with the incoming signals at the mixers 440, 445,
regardless of whether the incoming bits are odd bounce wavelets or
even bounce wavelets. Since this is a completely coherent process,
all of these signals will add coherently at the beginning of the
path and will all be integrated as either all positive or all
negative depending on whether the current symbol happens to be
positive or negative.
As a result, the output of the main path ADC 780 should have a one
polarity if the incoming symbol was a positive and the opposite
polarity if the incoming bit was negative. The polarity
corresponding to the output value will depend upon the polarity
chosen for the weight values W.sub.1 to W.sub.N. But regardless,
the output of the main ADC 780 is effectively a sign bit used for
detection of the transmitted symbol. Preferably, when symbols are
bi-phase modulated, the main ADC is arranged so that the least
significant bit straddles a value of exactly zero, so that a
decision is always made as to whether the symbol is detected as
positive or negative.
Bit Values
Because reflections are properly accounted for, the output of the
main ADC 780 will include sign and magnitude. These can be used for
multiple purposes. They can be used in soft decision forward error
correction decoders. They can also be fed into an equalizer circuit
to mitigate interference from previous symbols that have traveled
over longer multipath routes and arrive as interference, also known
as inter symbol interference (ISI). In this way the receiver device
720 can have an equalization process based on the decisions at the
main ADC 780.
Preferably, this is done in a post processing of the output data.
The equalizer can also be allowed to alter the weighting values
(W.sub.1-W.sub.N) used to weight the output of the wave-generators
750.sub.1-750.sub.N, 755.sub.1-755.sub.N With the proper time
shifts and weight adjustments the equalizer performance can be
enhanced. In other words, rather than using one of the
wave-generators 750.sub.1-750.sub.N, 755.sub.1-755.sub.N to
maximally correlate the current symbol coming in, the receiving
device 720 can use them to minimize the interference from a
previous symbols. The wave-generator essentially reverses the
process it would use to correlate the current bit so as to
destructively interfere with the the undesirable symbols in the
incoming signal.
In this manner the receiving device 720 could also cancel out
individual users through the use of codes. For example, if a first
user has a first code, and a second user has a second code, the
receiver device 720 could feed back the code for the second user so
that one of wave-generators 750.sub.1-750.sub.N,
755.sub.1-755.sub.N could cancel out the second user's signal when
the receiving device 720 is looking for signals from the first
user. This can effectively solve a multi-user detection problem.
Thus multiple users can transmit and collide with each other in the
air, but the receiver can null out all of the users but the desired
user (e.g., in a two user system it can null the second user when
it wants to receive from the first user and can null the first user
when it wants to receive from the second user.)
By using one channel to listen to a desired user, and using a
second channel to listen for a second user, the outputs of the two
channels can be evaluated by a post processor to improve the
performance of both channels. Detections from the first user can
help the device 720 to cancel the first user's signal out from the
second user's signal, and the detection from the second user can
help the device 720 to cancel the second user's signal from the
first user's signal, etc. In this way the receiving device 720 can
effectively demodulate signals from all of the user's
simultaneously, yet with much better results than a single channel
could achieve.
Then, because the receiving device 720 knows the signals for all
users (e.g., the first and second users in the above example), it
can feed each back to the others to cancel them out for the user
for whom interference need be eliminated. This process can
significantly help increase the throughput of the device 720
because it can use the channel more effectively.
The preferred embodiment shown in FIG. 7 shows one way to process
the signals using a reduced number of transistors and with reduced
power consumption, as compared with the embodiment shown in FIG. 4.
This is because the receiver 720 in FIG. 7 has the same number of
wave-generators and weighting mixers as the receiver 420 in FIG. 4,
but need only implement two amplifier-integrator-ADC paths.
Second Raking Receiver
FIG. 9 is a block diagram showing a raking receiver according to a
preferred embodiment of the present invention. As shown in FIG. 9,
the rake receiver 900 includes an antenna 925, a front end 910, a
first splitter 915, a second splitter 920, a main summer 930, an
error summer 935, a main mixer 940, an error mixer 945, first
through N.sup.th main wavelet forming networks (wave-generators)
950.sub.1-950.sub.N, first through N h error wave-generators
955.sub.1-955.sub.N, a main amplifier chain 960, an error amplifier
chain 965, a main integrate-hold-dump (IHD) integrator 970, an
error IHD integrator 975, a main analog-to-digital converter (ADC)
980, an error ADC 985, first through N.sup.th main weighting mixers
990.sub.1-990.sub.N, and first through N.sup.th error weighting
mixers 995.sub.1-995.sub.N. N can be any integer value greater than
1, and is preferably between 2 and 32.
The first splitter 915 operates to split the incoming signal into
multiple raking channels R.sub.1-R.sub.M, which will each be split
by the second splitter 920 into an error and a main path. By way of
example, FIG. 9 discloses the operation of the rake receiver on the
first rake channel R.sub.1. The other channels are like the first.
This is effectively a cross between FIG. 4 and FIG. 7, where each
channel in FIG. 4 is now comprised of a channel as shown in FIG. 7,
to provide greater flexibility in combining the outputs for
multiple uses--i.e. Rake, multi-user detection, and
equalization.
In the main path of the first rake channel R.sub.1, the rake
receiver 900 will multiply the incoming signal from the first rake
channel R.sub.1 with a plurality of wavelets generated by the first
through N.sup.th main wave-generators 950.sub.1-950.sub.N and
weighted by the first through N.sup.th weighting values
W.sub.1-W.sub.N through the first through N.sup.th main weighting
mixers 990.sub.1-990.sub.N. In this way the first rake channel
R.sub.1 correlates a waveform that is a combination of all the
weighted wavelets from the first through N.sup.th main
wave-generators 950.sub.1-950.sub.N.
This correlated signal will then be amplified at the main amplifier
chain 960 and then integrated at the maim IHD integrator 970. As
noted above with respect to FIG. 4A, the main IHD integrator 970
preferably has three integrators so that one can integrate the
current main signal, one can hold the most recent integration to
allow it to be output, and one can be dumping (or resetting) the
second most recent input value. The main IHD integrator 970
preferably has a switch that goes between the three integrators,
cycling through the pattern of integrate, output, and dump for
each. In an alternate embodiment is to replace the integrators with
a low-pass filter.
The output of the main IHD integrator 970 is then passed to the
main ADC 980, which converts the analog signal from the IHD
integrator 970 to a digital value that can be used by later
circuitry.
The error channel preferably operates in a manner essentially
identical to the main channel. It has first through N.sup.th error
wave-generators 955.sub.1-955.sub.N, which are weighted by the
first through N.sup.th weighting values W.sub.1-W.sub.N through the
first through N.sup.th error weighting mixers 995.sub.1-995.sub.N.
Although in this embodiment the weighting values W.sub.1-W.sub.N
provided to the first through N.sup.th error weighting mixers
995.sub.1-995.sub.N are the same weighting values W.sub.1-W.sub.N
provided to the first through N.sup.th main weighting mixers
990.sub.1-990.sub.N, in alternate embodiments they need not be. The
error channel can also be used to support M-ary PSK, or M-ary
QAM.
The error signal will then be passed through the error amplifier
chain 965, the error SDH integrator 975, and the error ADC 985, in
a manner analogous to the main channel described above.
In this preferred embodiment the main ADC 980 performs the bi-phase
demodulation for the raking receiver 900, i.e., the sign of the
output of the main ADC 980 determines the sign of the input bit. In
alternate embodiments, however, other kinds of bit detectors could
be used. For example, the device could use an amplitude modulation
detector. Or the entire channel could be made complex so that
instead of it being a single channel, it could have real and
imaginary components. In this case, the incoming signal would have
to be processed as a complex waveform.
In some preferred embodiments, the main wave-generators
950.sub.1-950.sub.N and the error wave-generators
955.sub.1-955.sub.N are provided as the same circuitry, but with
different outputs. FIGS. 10A and 10B are block diagrams of combined
wave-generator and weighting circuits according to preferred
embodiments of the present invention. As shown in FIG. 10A, one
embodiment of a combined wave-generator and weighting circuit 1000a
includes a simple wave-generator 1010, a first delay 1020, a second
delay 1030, a subtractor circuit 1040, and a weighting mixer
1050.
The simple wave-generator 1010 creates a wavelet. This wavelet is
then put through the first and second delays 1020 and 1030 to
provide an early value (before either delay 1020, 1030), a middle
value (after the first delay 1020, but before the second delay
1030), and a late value (after both delays 1020, 1030).
The subtractor circuit 1040 subtracts the early value from the late
value to get the wavelet for the error channel. The middle value is
used as the wavelet for the main channel.
The weighting mixer 1050 applies the weight W to the early value of
the wavelet before it is sent through the first or second delay
1020, 1030, and before the subtractor circuit 1040 receives the
non-delayed wavelet.
Thus, the combined wave-generator 1000 of FIG. 10 produces the
weighted wavelets used for both the main channel and the error
channel (i.e., it serves as main wave-generator 950.sub.1-950.sub.N
and error wave-generator 955.sub.1-955.sub.N, as well as main and
main weighting mixers main weighting mixers 990.sub.1-990.sub.N and
error weighting mixers error weighting mixers
995.sub.1-995.sub.N).
As shown in FIG. 10B, another embodiment of a combined
wave-generator and weighting circuit 1000a includes a simple
wave-generator 1010, a first delay 1020, a second delay 1030, a
subtractor circuit 1040, a main weighting mixer 1060, and an error
weighting mixer 1065.
The simple wave-generator 1010, first delay 1020, second delay
1030, and subtractor circuit 1040 operate as described with respect
to FIG. 10A.
The main weighting mixer 1060 applies the main weight W.sub.a to
the middle value wavelet (i.e., the wavelet after it has passed
through the first delay 1020, but before it has passed through the
second delay 1030). The error weighting mixer 1065 applies the
error weight W.sub.b to the output of the subtractor circuit
1040.
Preferably the main weight W.sub.a and the error weight W.sub.b are
the same, but in alternate embodiments they may vary by a small
amount according to the dynamic range of the circuits being
driven.
Thus, the combined wave-generator and weighting circuit 1000a,
1000b of FIGS. 10A and 10B produces the weighted wavelets used for
both the main channel and the error channel (i.e., it serves as
main wave-generator 950.sub.1-950.sub.N and error wave-generator
955.sub.1-955.sub.N, as well as main and main weighting mixers main
weighting mixers 990.sub.1-990.sub.N and error weighting mixers
error weighting mixers 995.sub.1-995.sub.N).
FIG. 11 is a top-level block diagram of a raking system according
to a preferred embodiment of the present invention. As shown in
FIG. 11, the raking system 1100 includes an antenna 1110, a front
end 1120, and a raking receiver 1130. The raking receiver 1130
includes a main path multi-weighted error wave-generator 1140, an
error path multi-weighted error wave-generator 1145, a main path
mixer 1150, an error path mixer 1155, and a main path signal
processing block 1160, and an error path signal processing block
1165.
The antenna 1110 receives an incoming signal and passes it through
the front end 1120, which performs some signal processing and
splits the signal into a main and error path. The main
multi-weighted wave-generator 1140 produces a main correlation
wavelet for correlating the main signal, while the error
multi-weighted wave-generator 1145 produces an error correlation
wavelet for correlating the error signal. Preferably the main and
error wave-generators 1140 and 1145 each generate, weight, and sum
N separate initial wavelets to produce their respective correlation
wavelets (where N is an integer greater than one).
The main path and error path mixers 1150 and 1155 mix the incoming
signal from the front end 1120 with the relevant main or error
correlation wavelets from the main and error multi-weighted
wave-generators 1140 and 1145 to produce main and error correlation
signals. The signal processing block then performs signal
processing (e.g., integration and analog-to-digital conversion) to
produce a final correlation value.
FIG. 12 is a top-level block diagram of a raking system according
to another preferred embodiment of the present invention. As shown
in FIG. 12, the raking system 1200 includes an antenna 1110, a
front end 1120, and a raking receiver 1230. The raking receiver
1230 includes a main path mixer 1150, an error path mixer 1155, a
main path signal processing block 1160, an error path signal
processing block 1165, a wave-generator circuit 1270, a main summer
1280, and an error summer 1285.
FIG. 12 clearly shows that a single wave-generator (or array of
wave-generators) can be used to drive both the main and the error
channels. This wave-generator circuit 1270 could be implemented as
an array of the wave-generators as shown in FIG. 10A or 10B.
The antenna 1110, front end 1120, main path mixer 1150, error path
mixer 1155, main path signal processing block 1160, and error path
signal processing block 1165 operate as noted above with respect to
FIG. 11.
The wave-generator produces a plurality of weighted main wavelets
M.sub.1-M.sub.N and a plurality of error wavelets E.sub.1-E.sub.N.
The weighted main wavelets M.sub.1-M.sub.N are summed at the main
summer 1280 to produce a combined main wavelet M that is provided
to the main path mixer 1150. Similarly, the weighted error wavelets
E.sub.1-E.sub.N are summed at the error summer 1285 to produce a
combined error wavelet E that is provided to the error path mixer
1155.
Multiple Raking Receiver
FIG. 13 is a block diagram of a multiple rake receiver according to
a preferred embodiment of the present invention. As shown in FIG.
13, the multiple rake receiver includes an antenna 1310, a front
end 1320, a splitter 1330, and M Rake Receivers 1340.sub.1 to
1340.sub.M, In one embodiment the raking receivers 1340.sub.1 to
1340.sub.M are formed as the raking receiver 1230 in FIG. 12.
The antenna 1310 receives a signal; the front end 1320 performs
initial signal processing on the signal; and the splitter 1330 then
splits the processed signal into M copies that are sent to the M
raking receivers 1340.sub.1 to 1340.sub.N, each of which produces a
main and an error signal (main-1, error-1, main-2, error-2, . . . ,
main-M, and error-M).
FIG. 14 is a block diagram of a controller and wave-generator
system for use in the multiple rake receiver of FIG. 13. As shown
in FIG. 14, the controller and wave-generator system 1400 includes
a controller 1410 and M wave-generator circuits 1420.sub.1 to
1420.sub.M.
The controller 1410 receives all of the main and error signals
(main-1, error-1, main-2, error-2, . . . , main-M, and error-M)
from each of the M raking receivers 1340.sub.1 to 1340.sub.N in
FIG. 13. Based on these signals, the controller 1410 produces
control signals for the M wave-generator circuits 1420.sub.1 to
1420.sub.M, which are formed in the M raking receivers 1340.sub.1
to 1340.sub.N, respectively. Each of the M wave-generator circuits
1420.sub.1 to 1420.sub.M requires N weight values (e.g.,
W1.sub.1-W1.sub.N for the first wave-generator circuit 1420.sub.1)
and N timing signals (e.g., T1.sub.1-T1.sub.N for the first
wave-generator circuit 1420.sub.1), assuming that the
wave-generator has N separate channels. Thus, the controller 1410
must generate (M.times.N) timing signals and (M.times.N) weight
values.
FIG. 15 is a flow chart of a raking method according to a preferred
embodiment of the present invention. As shown in FIG. 15, the
process begins when a receiver receives a plurality of signals.
(Step 1505) The receiver then performs amplification and filtering
on the signal (Step 1510), and splits that signal into error path
and main path signals. (Step 1515)
In the main path, the receiver generates a plurality of main
correlation wavelets (Step 1520); weights each of those main
correlation wavelets (Step 1525); sums all of those weighted main
wavelets together (Step 1530); and correlates that sum of weighted
main wavelets with the incoming signal that is coming from the
antennae. (Step 1535) After the incoming signal and the sum of
weighted main wavelets are correlated, the receiver detects and
outputs a main value. (Step 1540) That main value can then be fed
into a processor to control a wave-generator used to perform steps
1520, 1525, 1550, and 1555 for later incoming bits. (Step 1545)
In a preferred embodiment the main value is a bit value of a
current incoming bit. In this embodiment the bit value can be
detected by examining the polarity of the main correlation between
the incoming signal and the sum of weighted main wavelets. In
alternate embodiments, however, the main value can be determined in
other ways.
In the error path, the receiver generates a plurality of error
correlation wavelets (Step 1550); weights each of those error
correlation wavelets (Step 1555); sums all of those weighted error
wavelets together (Step 1560); and correlates that sum of weighted
error wavelets with the incoming signal that is coming from the
antennae. (Step 1565) After the incoming signal and the sum of
weighted error wavelets are correlated, the receiver detects and
outputs an error value. (Step 1570) That error value can then be
fed into a processor to control a wave-generator used to perform
steps 1520, 1525, 1550, and 1555 for later incoming bits. (Step
1575)
In alternate embodiments, steps 1510, 1545, and 1575 can be
eliminated.
In steps 1520 and 1550, the plurality of main wavelets can be
independently timed with respect to the plurality of error
wavelets, or the two sets of wavelets can be timed in
synchronization.
In steps 1525 and 1555, the weights of corresponding main and error
wavelets can be the same or they can be varied.
In step 1535, the correlation of the sum of weighted main wavelets
with the incoming signal includes mixing the sum of weighted main
wavelets with the main path signal that comes from the antennae and
then integrating that resulting signal over the bit interval using
an integrator. In alternate embodiments a sample-hold could be used
in place of an integrator.
Similarly, in step 1565, the correlation of the sum of weighted
error wavelets with the incoming signal includes mixing the sum of
weighted error wavelets with the main path signal that comes from
the antennae and then integrating that resulting signal over the
bit interval using an integrator. In alternate embodiments a
sample-hold could be used in place of an integrator.
In step 1570, eight-bit analog-to-digital converter may perform the
detection. The output main value could be one bit, which is just a
sign bit. However, in alternate embodiments the output main value
could be K bits, an amplitude modulation (AM) detector, or any such
signal.
In alternate embodiments the multiple channels could be complex so
they each have real and imaginary portions. This would require any
detection circuits be quadrature phase. In this case they could be
M-ary phase, or a combination of AM and phase. Any suitable
detectors could be used, however.
Multiple Wavelets for Bits
In each of the wave-generators disclosed above, multiple wavelets
can be used for a single bit. In this case a wavelet code would be
transmitted for each bit. In this case a receiving antenna would
receive a sequence of wavelets that matches that code for each bit.
The wave-generators in the receiver would then generate a series of
wavelets replicating that code sequence instead of just a single
wavelet.
Feedback
The various disclosed systems and methods can also use feedback.
For example, a processor can have available to it a string of bits
(i.e. a history of bits) starting at the current bit and going back
to a certain number of previous bits. The processor can then use
this string of bits to determine the proper weighting values to
send to the wave-generators in the receiver.
And if it chooses the proper weighting values, the receiver can
cancel out other users. In the alternative, the receiver could
cancel out one bit that falls on top of another bit so that even
though the number of multi-path components being received was high,
and many incoming signals were interfering with each other, the
processor could set the weighting values to cancel out the bits
that it is not looking for.
Also, just as it can cancel out undesired bits, the receiver may be
able to cancel out undesired users in a multiple user system. If a
first user employs a first code and a second user employs a second
code, a system that has multiple weighted wave-generators can set
some of the wave-generators to the first code, set some of the
wave-generators to the second code, reverse the polarity of the
weighting values for one of the users to null their incoming
signal. (This can be done to cancel out any number of undesired
users, provided that the receiver has sufficient wave-generators.
Thus, for example, when the receiver is trying to receive signals
from the first user, it can cancel out signals from the second
user, and when it is trying to receive signals from the second
user, it can cancel out signals from the first user.
Obviously, numerous modifications and variations of the present
invention are possible in light of the above teachings. It is
therefore to be understood that within the scope of the appended
claims, the invention may be practiced otherwise than as
specifically described herein.
* * * * *